damric

The beginning of the video was realistic enough to suck me in, but once he took the cooler off, well... The GPU socket though, my M2000 has something that looks like a damn AMD CPU socket. I didn't try to remove it, but clearly it's modular design. I appreciate the amount of work that they put into that video, and it did make me laugh.

You'll be missed. I can't believe I missed this post until now. Who is going to fill your shoes?

I was wondering why I was only getting 95% GPU utilization. Well the Ryzen 1600 had a single thread pegged to 100%, so I guessed it was bottlenecking somehow since it was running at default clocks and JEDEC RAM speed. I bumped CPU core to 4000MHz and the RAM up to 3200MT/s and now it's at 100% usage. Who knew folding could get CPU bottlenecked? 5% more utilization means a lot when folding 24/7.

Oh yeah that's always on my mind. I had it running for the last few weeks alongside my Vega64 in my main rig. Now it's moved to it's own board. I keep feeling the slot and the PCB on the board and the card and they are cool. I do have fans pointed on both front and back of the card to help. We'll see if this here B450 board can take it, or if it blows something, welds itself to the slot, or goes up in flames. I'm seriously playing with fire on this one, but you know that's how I roll. I can't wait to see the score in 24 hours. My first for the month's competition was completed by the W4100. As soon as it was done I messaged BWG and he changed it out on the score card. It will be interesting to see either it's destruction or a truly massive amount of ETF points. I need to set up a ring camera so we can all spectate as this speeding train goes off the rails.

Yup I had to retire that W4100. Yes it OC'd almost 100%, and yes it had a crazy multiplier, but it was so damn slow it was barely completing WUs by the deadline and missing out on quick return bonuses. Now I unveil the Quadro M2000. It always feel appropriate to put AMD fans on Nvidia product This could easily become another damric crash and burn story. I edited the firmware on the M2000 to bypass the clocks, but also the power limit was greatly increased. If you look closely at GPU-Z you can see that while it averages ~65W while folding, there are a * load of spikes up to 120W, and this card is only PCIE slot powered

Well I tried modding and flashing Michele's GTX 750 BIOS again. It failed to flash, no matter what version of NVFlash or KBT or MBT I used. It always says sanity check failed. I'm thinking EVGA put some sort of lock on it. I'm not too upset about it, it's already like 40% OC

I was looking at LARS again, and that entry for the GTX 750 Ti is very skewed. It's got it faster than cards like GTX 950, which means that whoever fed the last 500 submissions was juiced to the gills. I suggest switching to another card for ETF. then running that GTX 750 Ti on another key and feed it to LARS at stock clocks. Not the factory overclocks (if any), use the stock clocks for GTX 750 Ti according to techpowerup database which is base 1020MHz, 1085 Boost. Do 500 subs, and hopefully it fixes it since it only counts the last 500 as far as I can see. That should bring the stock PPD back in line to what it should be, which is somewhere between a stock GTX 750 (non-Ti) and a GTX 950, and fix your handicap multiplier so you can fold again on it next month. I hate beating you guys if it's not fair, and that * ain't fair.

Great job team! Since this was our debut month, I couldn't be more proud or happier about our strong performance. @mmcdonald12 I don't think you had any down time this month, which made a huge difference. And I am so glad we finally got @JosephStewartPE sorted out. I'm sorry that W2100 was a bust. In theory with 100% it should have mauled the competition same as my W4100, but it turned out that on the battlefield it was just too slow to complete units reliably. Thank you for your patience and perseverance. You really pushed us in the end when we needed it. I'm heavily considering retiring the W4100 for the same reason. Even with 100% OC it's just slow and can't even get 1 WU done per day.

When I join the metaverse I would like to buy an NFT picture of Zuckerberg jacking off to an NFT image of Zuckerberg.

Damn necros, now I'll be laying awake all night not knowing if he could hit 5GHz

I had no idea about any of that. Where did you find the good ones. All I ever saw was the regular one at TPU.

I had to use that flashy thing that bypasses the checks that you linked back in this thread somewhere. It's the only thing that would work for me. Now I'll try that on that GTX 750 and see if I can get anything else out of it with some added voltage.

Ok well it looks like it is stable for folding at over 1500MHz. I was able to run 3D benches at over 1550MHz. Anything over that crashes at this voltage. I'm not going to bump voltage as that will probably severely increase the power draw. I'm happy with these statistics I pulled from GPU-Z logs. Even with the huge overclock, it's averaging only 60W and has occasional peaks over 100W. I'm ok with that. Next month it will fold non-stop for ETF, so we'll see if there is a longevity issue with the slot power, but I bet it will be fine. It's still on an air cooler. I have a better air cooler coming, and I do have a water block if I really want to get the temperature down, but as you can see it runs cool already. While this M2000 card itself is too weak to be significant, there are greater implications now if someone wanted to overclock larger Maxwell Quadros now that we have a working theory. I see the M4000s (similar to GTX 970 I think) are going for less than $200, which is still a very good 1080p card for a desperate gamer in desperate times. GPU-Z Sensor Log-folding1506.txt GPU-Z Sensor Log-heaven-1557.txt

So far I'm at 1455MHz without issue. Of course HWBOT is down... Gonna let it fold for the rest of the evening and test some more tomorrow.

What I did was I dissected the ASUS GTX 950 card's BIOS which has no PCIE connector. I copied over much of the voltages and clocks, but kept my own elevated power limits. Thus it's still recognized as a Quadro M2000, but behaving more like the Geforce cousin. I'm going to let it fold overnight at the 1350MHz and tweak it some more tomorrow.

After some roadblocks, I cracked the locks on this M2000. I have a slight OC on it now. Now I need to see how much headroom this baby actually has, and how much power this PCIE slot can handle. So far with a small OC it's pulling about 55W and I saw a momentary spike of 88W. Temps nice and frosty now that I put a different air cooler on there.

messaged you in discord. your inbox here isn't working

You're all about to be overrun by 1D10T5. Joe should pop his 5th unit soon and Michele finally requalified for the quick return bonus a couple days ago.

Yeah I do it in bursts when I get something interesting I'll run a lot of stuff to submit, then I won't submit for a while. I'll have to run some benches with this 2700X and this Quadro M2000 as soon as it's done its 500 subs to LARS. I do miss you complaining about ETF though. I learned a lot about folding from you and the others, enough that I felt comfortable enough to try making my own team. Mixed results on that but there was a moment yesterday morning when we were in 2nd place

What gave you the itch to start benching and submitting to hwbot? I see you have been busy hard at it since you joined a few months ago.

People running them at max overclock while in dark mode skewed the PPD? But it's weird that there are even 2 entries for the same GPU.

I'd hate to help you guys but you should check with @BWG and be sure that handicap multi is right on that GTX 750 Ti. It seems like it should be a lot higher, like much closer to the huge ass multi that Michele has with her GTX 750. the PPDs are like GeForce GTX 750 Ti Folding@Home PPD Performance - folding.lar.systems FOLDING.LAR.SYSTEMS F@H GeForce GTX 750 Ti performance as of 1/25/2022. Averages across all projects PPD:81,018 - Work Units Per... and GeForce GTX 750 Folding@Home PPD Performance - folding.lar.systems FOLDING.LAR.SYSTEMS F@H GeForce GTX 750 performance as of 1/25/2022. Averages across all projects PPD:69,207 - Work Units Per... So something looks off because your PPDs look right but your handicap multi seems too low? I didn't math it, just estimating.

The great thing about benching older cards on the 3Dmarks is that you get so many extra points for physics with the modern CPUs. It really pads the score. They should really change the rules to just the graphics score but if/until then, run those points up!

I have about 250 stock samples left then I can play around with the overclocking. I'm thinking around Monday or Tuesday it should be ready.

Yeah so many entries! I'll be experimenting them all. Seriously I better be pushing at least 1400MHz and I'm hoping for closer to 1600MHz, under water ofc. I'm ok with pulling 100W on the PCIE slot. I've seen some cards exceed even that, pulling up to 120W during furmark runs. Not a big deal if it explodes while folding, as long as I get some HWBOT points first. I'm about to open up the BIOS you made and snoop around in there.